Ethernet vs USB? Which is better for communicating and automating test equipment and devices? Is it easy to implement Ethernet on an embedded system?
What are the common Design Rule Check errors that Parker and Stephen see as Contract Manufacturers? Are these DRC errors the ones that you run into?
On this episode, Parker sifts through marketing gimmicks for component ratings and Stephen discusses dual rail power supply designs.
Figure 2: Stephen’s discrete DIP-8 package opamp
Figure 1: 3D Layout of the Super Simple Power Supply and FX Dev Board
Parker is an Electrical Engineer with backgrounds in Embedded System Design and Digital Signal Processing. He got his start in 2005 by hacking Nintendo consoles into portable gaming units. The following year he designed and produced an Atari 2600 video mod to allow the Atari to display a crisp, RF fuzz free picture on newer TVs. Over a thousand Atari video mods where produced by Parker from 2006 to 2011 and the mod is still made by other enthusiasts in the Atari community.
In 2006, Parker enrolled at The University of Texas at Austin as a Petroleum Engineer. After realizing electronics was his passion he switched majors in 2007 to Electrical and Computer Engineering. Following his previous background in making the Atari 2600 video mod, Parker decided to take more board layout classes and circuit design classes. Other areas of study include robotics, microcontroller theory and design, FPGA development with VHDL and Verilog, and image and signal processing with DSPs. In 2010, Parker won a Ti sponsored Launchpad programming and design contest that was held by the IEEE CS chapter at the University. Parker graduated with a BS in Electrical and Computer Engineering in the Spring of 2012.
In the Summer of 2012, Parker was hired on as an Electrical Engineer at Dynamic Perception to design and prototype new electronic products. Here, Parker learned about full product development cycles and honed his board layout skills. Seeing the difficulties in managing operations and FCC/CE compliance testing, Parker thought there had to be a better way for small electronic companies to get their product out in customer's hands.
Parker also runs the blog, longhornengineer.com, where he posts his personal projects, technical guides, and appnotes about board layout design and components.
Stephen Kraig began his electronics career by building musical oriented circuits in 2003. Stephen is an avid guitar player and, in his down time, manufactures audio electronics including guitar amplifiers, pedals, and pro audio gear. Stephen graduated with a BS in Electrical Engineering from Texas A&M University.
Special thanks to whixr over at Tymkrs for the intro and outro!
Hello, and welcome to the macro fad engineering podcast. I'm your host, Parker, Dolman.
And I'm Steven Craig.
So as we were talking about last week is we're going to do a kind of live discussion of our super simple power supply. Build, talking about how the digital and analog interface is going to work. Live design. Yeah, live design. I've actually done zero prep for this. I think Stephens got some schematics.
I pulled a schematic that has maybe 10 parts on it. So I just may be one step ahead of you. Just one tiny step just once
I can see it. Yeah, yeah. All right. He just handed me this guy. This is for some I've sent Oh, it's just the op amps and some resistors. Exactly.
So it really doesn't even tell us a whole lot. No, doesn't but at least gives us some, some where to start.
Okay, so let's say it's, I guess the first thing go for is let's talk about how do we control the voltage outputs of the device at the op amp? And so it's going to be what signal? Do we have to give the op amp from the digital side? To do our varying voltage?
Right? Well, okay, so the digital side is going to have to create some kind of analog signal to pass off to the power op amps. So now we're just trying to come up with, I guess, a scheme, you know, is it? Are we doing zero to five volts? Are we doing negative 10? To positive 10? Or what? We just don't know yet?
Yeah. So what's gonna be the best way for giving the OP? What does the op amp want to do are? Our output
will? Okay, so let's, let's take a step back and look at this, the op amp design it basically, it's an inverting op amp configuration, just a really massive op amp on the output. Yep, it's powered by positive negative 35 volts. So you can give it bipolar signal. And it's just AC DC output. Okay. And so there's nothing particularly special there, we can apply game to it. We don't necessarily have to, we could always feed it a full scale signal and have a gain of one. And so, you know, if you put 35 volts in, it would spit 35 volts out,
that's gonna be a little hard to do, right?
Or, you know, if you put a let's say, we had a signal that we applied 10 times gain, you could put 3.5 volts and get 35. Yeah,
I think something that because on the digital end, we're gonna have 3.3 volt max. Okay, so a gain of 3.3 times to 35 volts. So it's gonna be a little bit above 10.
Right? Well, okay, something we got to remember this power supply. Each output is bipolar. It can swing the full negative 35 to positive 35.
So we have to go. I know it sounds more like we need a discreet, like a DAC.
Yeah, that's, that's probably going to serve us but like
a fast digital analog converter. Yeah. And then just talk surly to that. Mm hmm. And that way, we can actually just make a sine wave. If we're doing AC and that kind of stuff.
Yeah, that's probably the best. I think, what's what the big trip up is gonna be? Is honestly not not getting voltage out. But But where's our zero? Do you get what I mean by the Yeah, so So in other words, let's say we had a DAC that had an output of zero to five volts, just as an example. Yep, you could put your zero right in the middle at two and a half. And if you so if you wrote two and a half to the deck, then you get zero on your output. Anything below two and a half would be negative voltage, anything above would be positive. That's just an example. We could we could run something like that. Or we could actually apply an offset in analog. And so you could still write zero to five volt and then we apply a two and a half volts offset.
Now we can we can do the other we use like a 16 bit DAC, which should be plenty resolution. We could totally do. Basically halfway is the halfway is the zero volt. Sure.
So actually, yeah, that sounds like a good idea. When it comes to this really big The output up here, I think from an analog perspective, it's probably a good idea not to put a lot of gain on the output, correct. So that we can maintain as much stability as possible and lower the noise. Right? So I'm almost of the opinion right now that we should leave the output at a gain of one, the the, the big monster output, just have it at basically a throughput. And we can have an A an op amp in front of that, that applies all the game.
Oh, like a high precision game and prefer Yeah. So
so if you think about it, if you if, I guess, since we're just, you know, we don't have the ability to write down things right now. Think about it from from left to right. In your mind, you have the digital control that talks to a deck. That deck would then have it I'm sorry, add not a deck. Oh, yeah, you're right. No, I'm sorry, other way around my bed, a DD a DD, and then that would send off to an op amp that has a some gain to it, that op amp would then go to the output op amp. Okay. And that would allow us to achieve the full full scale range. Yeah, that
sounds good. Okay. Um, digital side of that, it would be really cool to basically, you would feedback is you would read your output from your DAC and the output of the whole system? Well, so you can error, correct? Yes, you can error correct. In case there's any temperature drift?
Yeah, yeah. Actually, he's
gonna happen no matter how well designed we make this thing, there's gonna be some drift.
Well, and and the tolerance of all the components is just, it's not gonna be spot on. Yeah. So it almost sounds like, you could have like a calibration routine, where it could actually, you know, go step through all the values and kind of almost make a lookup table forever. Yeah,
you can probably auto build its lookup table. Yeah, what its gain actually is.
That's, that's actually pretty cool. Yeah. And, well, okay. So if you think about that, if we set if we say that we're using a DAC with a five volt reference voltage, therefore, our center point or zero point would be two and a half volts, we would need two and a half volt to go to be gained up to 35 volts, right? Yes. So I got my calculator here.
No, 2.5 volts is zero.
Right? But but you swing 2.5 volts, up and down from 2.0. Okay, yeah, got it. Right. So if you divide 35 by 2.5, we're gonna need a gain of 14. Okay. Which that that's not that. That's not a huge gain. That's probably not going to add a ton of noise if we if we choose a good op amp. Sounds sounds fairly reasonable. Yeah,
that sounds pretty good. Cool. And now, so that's voltage control. So we have feet how we want to do the feedback. Yeah, voltage. Now, current limiting and current control. Right. So how does that op I know that op amp, the OPA 154541541. Has current limiting built in? That's great. How do we control that though? Or is it set? And that's it? Well, okay, so,
I, the the RPM kind of has, Oh, crap, current limiting, okay. Where it basically has a its output eggs actually fed across a low value resistor, such that it reads across a bass junction inside the op amp, okay. And once it reaches point seven volts, or whatever that base junction is, then it just kind of shuts off. It basically ramps it down. So it's, it's a it's a, it's not a really good current limit.
No, it's a, you just short circuited the whole thing. Yeah, you're done. messed up. Yeah.
So So what we could do is because we already have that resistor in play, we could actually just read the voltage off across that resistor, and read and have a feedback into our controller. Okay, yeah, that'll work. Yeah. So that'll at least tell us what the current is.
Yep. Yeah, that'll work. And we just in will just dial the voltage back for that, to keep the current at what needs to be. Yeah,
yeah, exactly. And that work? Because I guess, yeah, I don't think we're going to be able to actually physically control the resistance.
Yeah. And so the way because we'll be using the to measure basically the voltage outputs of the DAC and the op amp, we'll be using single ended ADCs. If I want to use a differential though, to measure that resistor.
Yes, yeah, you're right. Because because current can flow either direction.
Yeah. And well, it's not really either direction. It's just you get a much with different usually with a differential ADCs, you get a bigger your your resolution is a lot higher. Yeah, because you basically uses almost the same circuitry inside except the voltage differentials it's measuring is a lot smaller instead of zero to five volts. It's measuring like, you know, 50 millivolts Max difference.
Right, right. And it and it has really good common mode rejection. Yes. So so you get better noise performance out of a diff amp, as opposed to something else. And the resistor on the output of the op amp that I have right now on our board is only 300 million ohms. And I think if I can't remember the calculation off the top my head, but that's, I believe, that sets the current limiting at a few amps. So it's even going to be smaller than that in real power supply. Yep. So we're not I mean, we got a really low resistance here.
Cool. So I think that's all the things we need to control on the op amp side.
Yeah, yeah. I mean, it sounds like a scheme that could work well.
Yeah. Now and you've been working on the the regulated power supply that powers the power supply.
Right, so they get these op amps. Since they have a 35 volt positive negative rail. They got to get that from
somewhere. Yeah, somewhere. It's got to be beefy. Because got to pump out power. Yeah, it's got to be very beefy. Yeah, cuz we're talking three of these op amps in parallel. per channel. Yeah, six of these guys.
Yeah. And this is not something that you can, you know, pull out your 7805 regulator and put on this, this is a, it's got to be able to supply continuous 35 volt 10. Amp this this regulated power supply. So it's, it's gonna, it's effectively going to have to be a beefy pass through transistor style with a with an error amplifier.
Yeah. And you were designing this in multi sim earlier this week. That's right.
Yeah. So I've been playing around more multi sim. And to be honest, I'm loving. I'm loving every minute of it. It's working out really, really well. So I got a I have a regulated power supply. Simulated up right now both a positive and a negative 35. volt. And I, in the simulation stress test it and as of right now, it's holding steady at 35 volts and putting out a no 700 watts. Oh, that's cool. Yeah, it's, it's, it's absolutely beefy.
Yeah, you're also um, is that simulated the transformer stir?
Yes, absolutely. If you if you're doing if you're doing a, you know, a line voltage regulated supply, you got to have the transformer in there. And I went far enough that I'm not using an ideal transformer. I actually, from the transformer that we've selected in the datasheet, it gives numbers that you can calculate the coil resistance.
Yeah. And what they call the power loss factor.
Yeah, I think the power loss factor applies when you're talking about from the primary to the secondary. Okay, basically, the the the core has some kind of magnetic saturation, or basically a minimum current for the transformer to play its game effectively, let's just consider it that. And so at no load, the transformer actually draws like 12 watts. So it's a it's a beefy, beefy transformer. But But even with taking into consideration the resistance of the secondary coils, I'm still able to maintain the 35 volt regulation. Yeah,
those transformers are 16 pounds each.
Yeah. And we wanted the toroidal actually, donuts, yeah, good donut transformers, keep the noise down, make a little bit more efficient. And actually the manufacturer that we we've selected for it, they claim that these transformers can actually be stressed 20% higher than their rated values. So
and I think we're under the red values to
Yes, we we engineered a bit of fat in the design.
Yeah. We actually had two of these transformers as well in the sky, right? So it's 30 pounds 32 pounds of just transformers.
So and we had to, to get the ripple to from the rectifier to get it to an acceptable level. We had to put in 40,000 microfarads worth of capacitance.
Yeah, I've got actually been modeling up all the stuff that that Stephen needs in SketchUp to make sure all this stuff will actually fit along with all the cooling that we need, because I expect like a giant enormous 140 millimeter by 280 millimeter by 80 millimeters as a huge radiator with two huge fans on it. And I think in that guy can cool about 400 to 500 Watts, which is about a little bit over than what we need, which is
probably a good thing. Yeah,
it's gonna be a good thing. Yeah. But yeah, so the 3d layouts awesome. Pretty much done.
It's looking pretty good. Yeah.
And you actually going to have an article, right?
Yeah, yeah, we should be releasing an article tomorrow. Today's Thursday. So we released an article tomorrow on that.
Yeah. So it's actually going to come out the day that you listen to this podcast.
You Right, right. We record we record that the day before?
Yes. See, I've made that mistake before. Now you've made that mistake.
So we're even. Yeah, so we so we will have an article with it. And in the article, we're going to be talking about just some some bullet points on handling the the heat that we're generating in there and and you'll be able to see the simulations I got for the power supply. Yeah, it's
a gonna be a pretty good article, huh. Let's see. And then. So I think that's pretty much done on that end. Oh, yeah. The giant bridge rectifiers that you found? What's the package that you remember what the package name for those guys are?
Oh, man,
they're big. They're like two inches by two inches. By like a quarter inch thick giant bridge rectifiers that we can basically bolt right to the copper block. Yeah,
they're really great. They have excellent thermal bonding internal. Yeah. So honestly, you can't, can't ever neglect cooling down your rectifier.
Those things get hot. Yeah. Especially with how much amperage when we pull them through the guys. That's right. I think they're probably a little overkill. I think they're also like $34 like that apiece. Well, the thing is,
if we're running at our max load, our current pulses that come in, they're about eight milliseconds long. Each pulse and they're 45 amps. So we have to, these are a little bit overkill, but we still needed some big defiers,
also, speaking of the pulses is we're actually I am still trying to debate whether or not to actively cool our capacitors because they're going to get a little warm, because of how much current Sergi knows is going to happen through them. Yeah. And I've actually been thinking about taking a block of aluminum and milling out the, I think the 1.4 millimeter or 1.4 inch holes, and then basically slathering them in thermal grease and sliding them into it. Just so that has a heatsink basically for the cap. Yeah, it might be bit overkill. This whole thing is a bit opening is overkill. I've been working on that. Your FX dev board a bit? Yeah. Got the enclosure, mostly designed. The 3d modeling is these done, I got basically draw up a mechanical diagram that I can send over to the machine shop. And then they can build it. They can look at SketchUp and go. Oh, yeah, we can build that to bed. Yeah, yeah. Basically give them what says it's going made stamped steel, basically draw it out how it is when they stamp it out, and then show them where they need to fold the corners. Yeah. And that should be about it.
Have you? Have you actually. So the way you drew it up in Google SketchUp. Have you ever played with SolidWorks at all? I've never used SolidWorks. SolidWorks has a really cool function where you can draw it all up. And then you press a button and it'll automatically fold it out. Oh, that's cool. And it just gives you a DX F.
It's always SketchUp had that. Already done? Yeah. And then you showed me earlier this week, I think was on Monday. You built a discreet op amp?
Yeah, yeah. So I found a on my it's not my design. I'm I got a preference. It's by saying that on the website, DIY stompboxes. They have an a thread going on about a discreet op amp that has some pretty cool applications in in the audio world, specifically the Guitar World. And I saw that as like, oh my gosh, I got to make one of these. This is super cool. But I wanted it to be directly able to replace a an op amp like a tl 071 style op amp a single package. So I built a board that's the size of a dip eight package and did the whole op amp on that and you can literally just pull a pulling up inbound and plug this directly It's a drop in replacement, but it's all discrete.
It's a double sided assembly, right? Oh, yeah. Yeah. I thought it was pretty cool.
Yeah, yeah. Yeah. I I'm hoping to throw it in a stomp box that I've been building and see what it does wipe
out your your op amps for that discrete guy.
Yeah, yeah. And I want to see if that works. Well, I've kind of want to make a dual stack PCB that has two op amps in it. So you can replace a dual package, which are a bit more common in the single packages.
And actually, that goes on the topic of 505. timers. Yeah, and there's a couple discrete five or five timer designs out there. But the old, you know, big and usually for for educational uses. Yeah. But would it be possible to build a discrete five, five timer?
That's the same size of the debate package. Now, there's a lot more stuff in there than a normal op amp. You got a lot more transistors? Yeah. A lot of resistors. Yeah, no, capacitors?
No, no, it doesn't have any caps. That's good. Because cats can get big. Yes. You'd have to get really small on all your parts. Yeah. So
I did a little bit of research. And I found that the smallest, discreet BJT I found was the the two s si r 523. V one T two L. If you couldn't tell I was reading off a sheet of paper. Because I could not remember is that and it's a DFN style package, which is written off sheet of paper. VML 0806 Dash three. Which I had no idea what oh, I got now. The physical size is is point eight millimeter by point six millimeter by point 30806. Dash three gerbing. Yeah, yeah.
I like it. You're just about to say I don't know why they
just now I don't know what the VML is from? It's my Bye. Oh. I cannot remember the manufacturer.
Is it one of the big guys? No.
Yeah, doesn't matter. Anyways, I figured out why they need numbers the least.
But that's that's for all, it says effectively half a millimeter by half a millimeter. I mean, it's a little bit bigger, but that's tiny. And
it's a DFN style. So all the pads are underneath. So technically, you could pretty much button them up right next to each other. Yeah, your assembly house will probably hate you. But you can do that.
So are you gonna try to make a 555 I think I want to
try to build one. And that with those transistors, now it probably won't have the same. You know, if you drop it in a probably work differently, like timings will be off? Because I'm not gonna be able to use the exact same, you know, BJTs sure that they have in there? Well, I think we pretty good dexterous and might have to use your idea of double stacking yards to get more logic in there.
He I mean, once you start looking at how small a debate is, it gets. Yeah, I don't know. I mean, give it a shot. That sounds cool. Yeah, it does this code, just both NPN and PNP.
Um, I think they just have NP ns. But I think there's a one that's like a point eight by one millimeter. And you can get a PNP in that package. Okay. And same DFN style of things underneath. Yeah. And then, you know, use like, oh, 201 resistors
105.
We can build that. But I mean, the thing about Oh 105, at least from what I've seen is that you don't actually gain that much real estate. No, not really, because your pads actually aren't that much smaller than Oh, 201 pads. Yeah. And there are I don't think there are any. Oh, 105 caps yet. Very few Oh, two a once you only get like point one mics and a couple of small guys. Yeah.
Well, and we do oh two Oh ones pretty regularly, right?
Yeah. Yeah. Microphone symbols. Oh, yeah. Yeah. Yeah,
I mean, we do. Oh, 402 is all the time. Yeah. And we
do Oh, two ones. Yeah. Cool. Yeah. See, I'm gonna try working on that. Maybe next week. Okay. RFO section, rapid fire opinions. So, Microchip published a USB Mass Storage loader, which kind of works like a USB stick where you basically plug in your device, dump a hex file and it automatically loads it.
Hey, that's nifty.
Yeah, it's nifty. But this is stuff has been around for a long time. A lot of different platforms that use this kind of style. Yeah, we've actually used one network before like this. So how microchip does is you actually have two microcontrollers. One acts as the programming bridge. So when you dump it, it goes on to that chip as a mass storage. And then that guy pushes the bits over and programs the other pic.
That kind of sounds like cheating in a way.
Yeah, it is. Because there's actually a lot of these kind of mass loader kind of things actually work that way. Yeah. Now we have used St. St. Was it? STM? Yeah. STMS that have this book that has this stuff. But it's all in one chip. Mm. Yeah. And it's like, and, like, even like, you can jump Python scripts on these guys. Right? And just you just drop it onto it and it works. Well, yeah, right. Right. Right. So I think microchip is a little late to the game on this well,
but isn't isn't microchip aren't they trying to do this as a way for loading your hex directly onto the chip? Yeah, cuz I think you you open up your USB and then you drop your hex and and and it just auto programs basically. Correct. Okay. Yeah. But, but I but I don't think it works like the STM where you can load a script, and then just run that program. That's true. So, uh, you know, I don't know. Seems, seems kind of interesting. Yeah, it's,
it's one of those. If you already have your computer there, why can't you just run the programmer and get rid of one of the microcontrollers? Yeah, it
seems a little excessive. Yeah. I
don't really see the, I guess appeal to to that process. Yeah. Sounds expensive, hardware wise. And then, we're going to continue with ragging on microchip. I installed MP lab, I think for like the fourth time at work on different computer, and they changed the logos for the icons again. And I was like, I only noticed that first time because I haven't used pics a lot. And I asked Steven, I'm like, hey, they change the logo. And he's like, again. So apparently the done how many times have they changed the logos?
This is this is easily the fourth or fifth logo that I've seen, for sure. And and I've only been dealing with pics for probably five years
or so. Yeah, it's um, the new logo is kind of like a badge almost to Yeah, it's like a, like a sheriff badge look to it.
Oh, and what was wrong with their MP lab X logo? I mean, it was all like, futuristic. And
it kind of looked like 90s Extreme.
Yeah, it got it.
Because it was like MP lab X and X was ginormous. So what yay, nay? They look good. It's just why did they change it again?
Yeah. So none of the logos have been bad. It's just why can't there be consistent consistency?
Yeah, I guess I've never looked at an icon and be like, that's a bad looking icon. Yeah.
Also, also, this is that, you know, that's an interesting thing about IDs. When it comes down to programming, whatever you're playing around with, why do I need to keep downloading tons of IDE like what is changing that I need to download? It's literally I write my code, and I push it into my chip. I mean, I could do that with with stuff that's 10 years old.
Yeah. Like, well, there's bug fixes and that kind of stuff. Yeah, sure.
I mean, I understand that. But can it just be an update, as opposed to like, a whole brand new package?
Yeah, I see that. Yeah. I guess it depends on how well, np lab, they kind of fix that kind of I guess, because you had to download the compiler separately now. Yeah. And I always didn't like that.
Well, I think they have a, they have like a pay to play compiler. I think they have like a regular compiler. And then they have what they called the high C compiler, ah, and and they show that like, if you if you get their special compiler, then it uses 30% less RAM or something.
Oh, because the compiler, basically when you install NP lab, it says, Do you want to go to the microchip website and download your compilers now? It's like, Ah, I thought this had compilers. So you have to go to the website and download. You can download the free ones there. Yeah. But I've always liked basically like T eyes, their Code Composer Studio stuff where when you get the stuff, you get everything. It all works, you don't have to configure anything. Whereas with MP lab, you have to install the compiler and make sure it works and you get links to yeah and link it all up. Yeah. Annoying. reduce friction. Always good thing. Yeah. Okay, and then there were some interesting development for flexible circuits. So flexible circuits are really cool. We already have flexible PCBs. But the problem with them is, generally the metal you put in them isn't flexible, the copper. And so eventually over, you know, 10,000 cycles, your flex circuit wears out. Right? Which makes sense, which makes sense. You've basically got metal bending.
Hmm. And there's a maximum bend radius on any of that stuff. Yeah.
I wonder if they made that tool steel. You wouldn't get much flex, but it can flex forever.
You mean, the actual copper traces or tool steel traces? Who will still traces? That sounds like a giant pain? Probably.
It would flex for a while though. Same. So if you make a leaf springs out of trucks? Yeah. Researchers at the center of neuro? Was it prosthetics? Yeah. At the Oh God, the Ecoli Polytechnic. Federal de luz Sonny.
You took French in high school, didn't you? Yeah,
totally. Um, so they develop this new metal. Basically, it's a alloy of gallium and and gold. And so gallium at room temperature is is a liquid. Gold's not. And I guess they mix them into an alloy that's kind of liquidus and kind of not. And they basically made it so you can bond to silicon. And they can make flexible circuits. And they have a really cool video of basically a grid of LEDs. And it like expands up in a big bubble and everything's still working. It's really cool looking.
I guess that's the fifth state of matter. Kind of liquid kind of, yeah, kind of. Yeah. Well, that's, that's cool. I'm wondering what the applications for something like that. Like, where's that going to go?
You can do well, you can wear you can make it wearables that kind of stuff or prosthetics.
Flexible
screens? Yeah. Or just like, all balloons, like party balloons, Blinky party balloons.
You know, I'm sure there's loads and loads of research money going into party balloons. That's why they
you know, why they've done it was funny. Of all the things we said I betcha led party balloons, which sell the best and, and get the most amount of return money over like prosthetic limbs and all these other things. So that actually helps, you know, party balloons that actually help from end party balloons will probably make more money. You know what,
I'm gonna quit? I'm starting led party balloons.
So over the weekend, I was with my parents, okay. And my, my mom had this really weird iPhone case for iPhone six. Okay. And I was looking at I'm like, What is this thing it says Lumi on the back and has a button and you press it. And it has LEDs all the way around the front side of the case. That light up. When I asked my mom I'm like, What's this for? And she's like, it's for taking selfies.
Oh, no.
Gosh. And she's like, the Kardashians love it. Oh. And I'm like, Why did you like this?
They found a product that's worse than a selfie stick.
Yeah, worse than a selfie stick. Oh, my mom owns one of those two. Oh, gosh. Oh, this is obviously in a selfie cell phone case.
Oh my god,
it reminds me of those old timey like makeup.
Like vanity.
Vanity Mirror. Yeah, it's that kind of style except with lots of little LEDs.
Gotcha. That is why some engineer had to sit out there and actually design this thing Yeah. What a poor soul.
Yeah. And I was just thinking is like Blinky LEDs you know, I'm always a big fan of Blinky LEDs, but I think that's it's gone too far.
i They ruin though.
You know, I wonder. This isn't this, you know, as we design stuff We keep you know, adding in Blinky LEDs because that's kind of like the cool thing. Like, I'm actually looking at all the recording equipment and it's got bar graphs and meters and a lot of blinking stuff was really cool to look at. I wonder if there's ever going to be a movement? For none of that.
You know, like, you mean like an aesthetic design in electronics? Yeah. For the non Blinky.
Yeah, for the non Blinky LEDs.
Oh, it's called apples.
Oh, no, they got the apple glowy logo on the back of their lids.
That is, yeah, you're right. Because everyone needs to know that you are on an apple.
Yeah. But they do hide a lot of them. They laser little tiny micro drills into it, and they put an LED behind it. So you can't actually see the LED. Unless it's on. Okay, yeah, yeah. So the only led you can see is the one that lets everyone know you own an apple.
Okay, and the finish up the RFO section for this week is a, there was an internet poll. It's kind of not related to electrical engineering. But it's an interesting article anyways, it's soared forward of it. Yeah, it deals with the Internet. There was a poll to name that new $288 million research, Arctic vessel, a basic contest for the name. And the current winner of the contest is the RRS Boaty McBoatface. And the funny thing is, it wasn't actually the intern who originally came up with a name. It was a DJ, in Britain, tweeted about it and made like made fun of of the contest. And the internet ran with it.
Of course they did. Of course they did. The we should we should all learn from these kinds of mistakes that you cannot let the internet make decisions for.
Exactly. Never let the internet make any kind of decision. No. And if you do, you can make sure that one IP address per vote. Yeah, yeah, that's true. There's been a lot of instances where people just hammer
stuff anonymous will catch wind of it. And it's game over after that exactly.
But this goes into what happened early in URL later this week, I think was yesterday was Microsoft had this AI on Twitter, called Tay tweets. And it was supposed to is supposed to be a research project, so that they can make their automated calls better. Okay. And basically, he would learn from people talking to it, of how it should react, and how it should talk back. And it basically went from an innocent teenager to a Hitler loving sex robot in 24 hours. I mean, the stuff this thing was tweeting was insane. Yeah, it's almost like horrible stuff. Yeah, it's terrible. It's like one of those. They seriously let this thing loose in Devon monitored for 24 hours. So it looks like it goes
from a research project in in, you know, AI or whatever, to more of like a social experiment. Yeah.
Cuz like, you got to think like, once it started rolling, they would have either cut it or stopped putting that stuff into its database.
Yeah. Well, you know, they
just had a human sitting there and at least approving one tweets out.
Well, you never know, maybe it was part of the research thing where it's like, we're gonna let this go and see where it gets. That's true. No matter what. And it went really south.
Yeah, it was, it was pretty terrible. Ah, yeah. So yes. Never let the internet make any kind of decision for you know, and that's anything you take away from our podcasts that will be it.
And of all the podcasts, all of them even
future ones. Well, I think that's gonna wrap up this week's macro engineering podcast. I'm your host, Parker. And I'm Stephen. catch y'all next time. Take it easy.
What are the common Design Rule Check errors that Parker and Stephen see as Contract Manufacturers? Are these DRC errors the ones that you run into?
Ethernet vs USB? Which is better for communicating and automating test equipment and devices? Is it easy to implement Ethernet on an embedded system?
On this episode, Parker sifts through marketing gimmicks for component ratings and Stephen discusses dual rail power supply designs.